Restoration of the Bone Marrow Microenvironment Improves Hematopoietic Function in a Murine Model of Myelodysplastic Syndrome

While data suggest that, in the myelodysplastic syndromes (MDS), the bone marrow microenvironment (BMME) is abnormal, the lack of examination of the BMME in a robust in vivo model has limited progress in the understanding of reciprocal MDS-BMME interactions.  If microenvironmental defects contribute to disease progression, targeting the BM niche may offer an alternative approach for therapeutic benefit.  We sought to define the MDS BMME in a well-established transgenic murine model that recapitulates hallmark features of human MDS.  In this model, hematopoietic tissue-specific expression of the NUP98-HOXD13 (NHD13) fusion gene is driven by Vav regulatory elements, resulting in peripheral cytopenias by 16 weeks of age and mortality from transformation to leukemia at a median time of 11 months of age.  Mice were analyzed at 15-36 weeks of age, when the MDS phenotype is prominent in the absence of leukemia.  Flow cytometric quantification of BM stroma in 23-week old NHD13 mice showed a 6.5-fold increase in frequency of CD51+/Sca1- osteoblastic cells (OBC) compared to WT (p<0.05).  CD51+/Sca1+ multipotent stromal cells (MSC) and CD31+/Sca1+ endothelial cells were also significantly increased in NHD13 compared to WT mice.  This was not due to loss of hematopoietic cells in the marrow of NHD13 mice.  While an expansion of functional MSCs and osteoblastic cells could result in skeletal changes, micro CT imaging of the femora and tibiae of 20-week old NHD13 mice revealed no differences in skeletal parameters compared to WT mice.  These data suggest that the expanded osteolineage cells in NHD13 mice are not functional bone-forming cells.  While stromal populations were not altered in bone-associated cells of 23-week old NHD13 compared to WT mice, 36-week old NHD13 mice also showed increased bone-associated OBCs, MSCs, and endothelial cells.  Therefore, there are significant time-dependent shifts in critical stromal populations in this in vivo model of MDS, which may contribute to an abnormal BMME.  To determine if the MDS BMME contributes to hematopoietic failure, NHD13 BM (CD45.2) was transplanted with WT competitor BM (CD45.1) in a 1:1 ratio into lethally irradiated NHD13 or WT (CD45.2) recipients, thus exposing the same MDS hematopoietic cells to either MDS or WT microenvironments.  Using this transplantation paradigm, we previously reported improvement of hematopoiesis when NHD13 BM is exposed to a WT BMME.  Surprisingly, CD45.1+ WT competitor-derived cells exhibited myeloid skewing when transplanted into NHD13 recipients compared to WT recipients, suggesting that interaction of WT BM with an MDS BMME can induce myeloid skewing, a feature of the NHD13 model.   NHD13 BM was next transplanted non-competitively into lethally irradiated NHD13 and WT mice.  At 10 weeks post-transplant, WT recipients had a 2.5-fold increase in peripheral leukocytes (p<0.05), significant improvement of anemia, and significant mitigation of BM long term-HSC loss compared to NHD13 recipients, suggesting that WT BMME components can rescue hematopoietic function in MDS.  Together, our studies strongly suggest that a murine model recapitulates MDS microenvironmental abnormalities, and that exposure of MDS hematopoietic cells to a non-malignant microenvironment is sufficient to improve hematopoietic function.  Thus, improvement of the BM microenvironment represents a novel therapeutic strategy to ameliorate hematopoietic function in MDS.